Proceedings of the Nutrition Society (1992) 51, 71-79 71

PROCEEDINGS OF T H E NUTRITION SOCIETY

A Meeting of the Task Force on Overseas Members was held at the University of Edinburgh on
27-30 August I991

Symposium on
‘Nutrition and development’

Jamaican studies in nutrition and child development,

and their implications for national development
BY SALLY M. GRANTHAM-McGREGOR AND G E O R G E CUMPER
Tropical Metabolism Research Unit, University of the West Indies, Mona, Kingston, Jamaica

Protein-energy malnutrition (PEM) is diagnosed by comparing children’s anthropo-

metric measurements with reference values for their age and sex. It is estimated that 177
million children under 5 years of age in developing countries have weights-for-age below
two standard deviations of their expected value. Low weight-for-height (wasting) is less
common but low height-for-age (stunting) is estimated to affect 35% of children between
24 and 59 months of age (Grant, 1991). If PEM has a detrimental effect on children’s
mental development the implications are obviously enormous.
In the present paper, studies which have addressed this problem in Jamaica over the
last 15 years will be discussed. No attempt is made to review the international literature
which has been reviewed recently (Simeon & Grantham-McGregor, 1990). First those
studies concerning severe and mild to moderate PEM in young children will be discussed,
then studies conducted in school-aged children.

SEVERE PEM

The Wellcome classification for severe PEM will be used here (Lancet editorial, 1970).
Many studies have shown that children who suffered from severe PEM in early childhood
and subsequently return to poor environments usually have poorer mental development,
school achievement and more behaviour problems than their peers or siblings
(Grantham-McGregor, 1989). However, in spite of reasonably consistent findings, a
causal relationship cannot be inferred unequivocally, because of study design problems.
These include: the lack of adequate control for the confounding variable of poor social
background, siblings have usually been undernourished themselves and there is generally
no evidence of normal development before the episode of PEM. In addition the
malnourished children have usually had long hospital stays which may have contributed
to their poor development.
Acute stage of severe PEM. In an early study, in order to control for the effects of
hospitalization, the developmental levels of eighteen severely malnourished children

Downloaded from https://www.cambridge.org/core. IP address: 114.125.72.109, on 23 Dec 2021 at 09:28:48, subject to the Cambridge Core terms of use, available at
https://www.cambridge.org/core/terms. https://doi.org/10.1079/PNS19920012
 72 S . M . G R A N T H A M - M c G R E G O R A N D G . CVMPER

were compared with those of fifteen adequately nourished children who were also in
hospital and were suffering from other diseases (Grantham-McGregor et al. 1978).
The children had developmental assessments on the Griffiths’ test (1967), during
recovery in hospital. Both groups had low scores on admission, but the malnourished
group was markedly lower. Also both groups had particularly low scores in the hearing
and speech sub-scale. Both groups improved on recovery approximately the same
amount so that the malnourished group maintained their deficit. No previous study had a
control group of adequately nourished children in hospital and the low language scores
had been attributed to PEM. Also the improvement in development quotients (DQ) had
been attributed to the children’s improving nutritional status (Cravioto & Robles, 1965).
It is more likely, that both these manifestations had similar aetiologies in each group,
such as being in hospital and generally feeling unwell.
Rehabilitation of severely malnourished children. Few studies aimed at improving the
mental development of survivors of severe malnutrition have been reported. The only
two with control groups comprised short-term stimulation while the children were in
hospital, and these produced only transient benefits (McLaren et al. 1973; Cravioto &
Arrieta, 1979).
We conducted a long-term programme of psychosocial stimulation with severely
malnourished children (Grantham-McGregor et al. 1987). The intervention comprised
play in hospital, followed by home visiting for 3 years. The visitors were community
health aides, who demonstrated the use of home-made toys to the mothers and
encouraged them to play with their children between the visits.
The intervened malnourished group ( n 21) was compared with two other groups who
were patients in the same hospital 1 year previously, and had received standard medical
care only. These comprised a second severely malnourished group ( n 18) and an
adequately nourished one ( n 21) suffering from diseases other than malnutrition. All
children were tested throughout the study on the Griffiths’ (1967) mental development
scales and the Stanford Binet.
On admission to hospital both malnourished groups had similar developmental levels
(DQ) and were seriously behind the adequately nourished group. All groups improved in
hospital, the intervened group improving the most. Over the following 9 years the
non-intervened malnourished group’s DQ or intelligence quotients followed a trajectory
below that of the adequately nourished children, and they showed little sign of catching
up. This concurs with findings from retrospective case-control studies conducted
elsewhere. The intervened malnourished children showed gradual improvement and
caught up to the adequately nourished group in the first 2 years, after that they declined
slightly. At 6 years after intervention stopped, they had significantly higher school
achievement grades in spelling and reading but not in arithmetic than the non-intervened
malnourished children. They were significantly lower than the adequately nourished
group only in arithmetic (Fig. 1). The locomotor sub-scale was the only Griffiths’ (1967)
sub-scale in which the intervened children did not catch up with that of the adequately
nourished group at any stage. Also they failed to catch up in height-for-age and head
circumference-for-age. It is possible that if the children had been provided with
nutritional supplementation after leaving hospital, locomotor development may have
improved. The study, therefore, demonstrated that a low-cost modest intervention could
benefit these children’s development and that benefits lasted at least 6 years.

Downloaded from https://www.cambridge.org/core. IP address: 114.125.72.109, on 23 Dec 2021 at 09:28:48, subject to the Cambridge Core terms of use, available at
https://www.cambridge.org/core/terms. https://doi.org/10.1079/PNS19920012
 TASK F O R C E ON OVERSEAS MEMBERS 73

fi
T

... . ~ .. .
I
*
.

..
.. .. ..
....
* . . *

.. / . .:
-. .. /.*. .
Reading Spelling Arithmetic

Fig. 1. School grades, 6 years after intervention (long-term psychosocial stimulation programme), in three
groups of children. ( B ) , non-intervened malnourished (n 16); (El), intervened malnourished ( n 17); (U),
adequately nourished (n 20). Values are means with their standard errors represented by vertical bars.

One child in the non-intervened malnourished group was adopted by a middle class
family. He showed remarkable improvement in development, which suggests that with a
vastly enriched environment marked improvements can occur (Grantham-McGregor &
Buchanan, 1982). This is also suggested in other adoption studies of severely mal-
nourished children (Winick el al. 1975; Monckeberg, 1990).
Association between development and nutritional status. On admission to hospital, the
height-for-age, but not weight-for-height or presence of oedema, of the malnourished
children described previously, predicted their level of development 1 month after
returning home (Grantham-McGregor, 1982).
In a small study, children who had recently recovered from severe PEM were
compared with children, matched for age and sex, who had a similar degree of stunting
but had not suffered from a severe episode (Grantham-McGregor et al. 1989). Both
groups were given developmental assessments on the Griffiths’ (1967) test, and their
DQs were only slightly but not significantly different. Both groups’ DQs were signifi-
cantly lower than a third group of fifteen non-stunted children of the same age and from
similar areas of residence (Fig. 2). It was concluded that in this population, the poor
developmental levels usually found in severely malnourished children may be largely
explained by factors associated with stunting, rather than the acute episode.

S T U D I E S WITH M I L D T O M O D E R A T E U N D E R N U T R I T I O N

A survey of the nutritional status and developmental levels of 168 young children was
carried out in two poor Kingston neighbourhoods. Their DQs on the Griffiths’ (1967)
test were significantly related to their heights-for-age but not weights-for-height , when
several social background factors were controlled (Powell & Grantham-McGregor,
1985). The relationship between stunting and poor development is usually, but not

Downloaded from https://www.cambridge.org/core. IP address: 114.125.72.109, on 23 Dec 2021 at 09:28:48, subject to the Cambridge Core terms of use, available at
https://www.cambridge.org/core/terms. https://doi.org/10.1079/PNS19920012
 74 S . M. GRANTHAM-McGREGOR AND G . CUMPER

T
110

I
100

U
n

90

80

Severe Stunted Non-stunted

PEM

Fig. 2. Developmental quotients (DQ) of three groups of children: (W), recovered severely malnourished
( n 29): (El), group matched for age, sex and height (n 29); (O), non-stunted controls ( n 15). Values are means
with their standard errors represented by vertical bars. (From Grantham-McGregor ef al. 1989.)

invariably, stronger than that with wasting (Grantham-McGregor et al. 1990). The
mechanism is not clear. It may just reflect the duration of undernutrition as wasting tends
to be more transient than stunting.
The effects of nutritional supplementation and psychosocial stimulation in growth-
retarded children. Stunting is usually associated with poor development. However,
stunting is also associated with poor environments (Martorell, 1988) and it is difficult to
separate nutritional effects from those of poverty in purely observational studies.
To determine whether stunting was linked to poor development through nutritional
deficiencies we conducted a 2-year trial of nutritional supplementation (Grantham-
McGregor etal. 1 9 9 1 ~ )We
. also examined the role of stimulation. Stunted children (129)
aged between 9 and 24 months, were randomly assigned to four groups: control,
stimulated, supplemented, and combined supplementation and stimulation. Non-stunted
children (32) matched for age, sex and neighbourhood to each control child were also
examined.
The supplement comprised 1 kg milk-based formula/week, and the stimulation was
weekly home-visiting similar to a programme used previously (Grantham-McGregor
et al. 1987). The controls were visited every week in the same way as the supplemented
children in an attempt to control for any increased attention the supplemented children
may have received. However, this was not a true placebo, as it was considered unethical
to provide low-energy supplements to stunted children. Children were evaluated every 6
months on four sub-scales of the Griffiths’ (1967) mental development scales (locomotor,
hand and eye coordination, performance, hearing and speech) by testers blind to the
treatment of the children. The stunted children initially had significantly lower scores in

Downloaded from https://www.cambridge.org/core. IP address: 114.125.72.109, on 23 Dec 2021 at 09:28:48, subject to the Cambridge Core terms of use, available at
https://www.cambridge.org/core/terms. https://doi.org/10.1079/PNS19920012
 TASK F O R C E O N O V E R S E A S M E M B E R S 75

110 t

105

x
100

95

Enrolment 6 12 18 24
Age (months)

Fig. 3. Developmental quotients (DQ) of stunted groups adjusted for initial age and score, compared with
non-stunted group adjusted for age only (Grantham-McGregor et al. 1991). (0- - - -O), Both; (O---O),
stimulated; (A-- --A), supplemented; (O--+), control; (M), non-stunted.

all sub-scales than the non-stunted ones. The control group declined over the 2 years and
ended 14 DQ points below the non-stunted group. Stimulation benefited all sub-scale
scores, whereas supplementation benefited scores in the locomotor and performance
sub-scales, and the effect on the hearing and speech sub-scales approached significance
(P<0-08). The combined effects of supplementation and stimulation were additive and
this was the only group to catch up to the non-stunted groups. The mean DQ of the
groups are shown in Fig. 3.
This study fulfilled most of the requirements of a clinical trial and provides evidence
that stunted children’s development improves with supplementation. The implications
are that at least part of their deficit in development is directly attributed to poor
nutrition.
Studies in schoolchildren. Several studies have been conducted in schoolchildren. In a
small study, breakfast was given to one class of remedial children for one term. They
improved in arithmetic and school attendance, but not height or weight compared with
two other matched classes (Powell et al. 1983). We hypothesized that the mechanism was
the relief of short-term food deprivation. A much larger study was carried out in grade 5
children in five randomly chosen government primary schools in Kingston (Clarke et al.
1991). An association was found between the children’s height-for-age, haemoglobin
level, breakfast history and days without food in the home (N. W. Clarke, personal
communication) on the one hand and poor school achievement on the other. After
allowing for extensive social background variables in multivariate analyses, the nutrition
variables still contributed to the variance in the children’s school achievement levels.
In an experimental trial of missing breakfast (Simeon & Grantham-McGregor, 1989),
stunted and non-stunted, 9.5-ll-year-old children were compared. Stunted children’s
cognitive functions were detrimentally affected when they missed breakfast, but
non-stunted children’s were not. These findings have implications for school feeding
programmes in developing countries. They have generally been poorly evaluated

Downloaded from https://www.cambridge.org/core. IP address: 114.125.72.109, on 23 Dec 2021 at 09:28:48, subject to the Cambridge Core terms of use, available at
https://www.cambridge.org/core/terms. https://doi.org/10.1079/PNS19920012
 76 S. M . G R A N T H A M - M c G R E G O R A N D G . CUMPER

(Levinger, 1986), but may be more effective than evaluations in developed countries
have suggested. These findings are also important because they demonstrate an
interaction between the children’s underlying nutritional status and short-term food
deprivation. Other nutrient deficiencies such as iodine and iron could also interact with
PEM.
MECHANISM

The mechanism linking PEM to poor mental development is not established; however,
there are several hypotheses. A frequently cited one is that the children‘s activity and
exploration levels are reduced as a result of PEM and they subsequently acquire skills
slower than the normal rate (Levitsky, 1979).
We examined this later hypothesis in two studies. The behaviour of severely
malnourished children was observed while they were in hospital in the acute stage. They
were compared with adequately nourished children who were ill with other diseases
(Grantham-McGregor et al. 1991b). Time-scheduled observations were used. The
malnourished children were initially significantly less active, explored toys less, touched
fewer toys and used fewer types of exploratory behaviours. They were also more
apathetic, whereas the controls were more distressed. All differences disappeared on
recovery, except that the malnourished children still handled fewer toys. This latter
difference also disappeared when the children’s DQ were controlled. Initial behaviours
did not predict DQ levels on recovery within the malnourished group. The rapid return
to normal in behaviour, contrasted with the children’s persistently poor DQ.
In the supplementation study of stunted children described above (Grantham-
McGregor et al. 1991a), the children’s activity levels were observed on enrolment, using
time-scheduled observations in the homes (Meeks Gardner et al. 1990). The stunted
children were less active than the non-stunted children. Activity levels were significantly
related to the locomotor sub-scale scores of the Griffiths’ (1967) test but not to other
sub-scales (Grantham-McGregor et al. 1990). Further, activity levels did not predict
future change in development (J. M. Meeks Gardner, personal communication).
Although it is established that severely malnourished and young stunted children have
reduced activity levels, the studies failed to confirm a causal link between reduced
activity and poor development. The role of activity in poor development, therefore,
remains questionable.

I M P L I C A T I O N S FOR I N D I V I D U A L A N D N A T I O N A L D E V E L O P M E N T

Though the interpretation of the Jamaican studies is limited by their relatively small scale
(reflecting the limited research resources available), they confirm that meaningful
nutritional intervention is possible in the Jamaican context, and in particular that
nutritional status is associated with individual developmental improvement and edu-
cational success or failure. This is important from a policy point of view because of the
demonstrated or suggested associations between education and individual or national
development (Fig. 4).
Published studies have demonstrated, or at least plausibly proposed, that education
influences the level of economic productivity in some occupations, and promotes modern
attitudes and values (Lewin et al. 1982; Schiefelbein, 1983). Education also improves the
development of cognitive skills in children (Wagner, 1974; Stevenson et al. 1978).

Downloaded from https://www.cambridge.org/core. IP address: 114.125.72.109, on 23 Dec 2021 at 09:28:48, subject to the Cambridge Core terms of use, available at
https://www.cambridge.org/core/terms. https://doi.org/10.1079/PNS19920012
 TASK F O R C E O N O V E R S E A S M E M B E R S 77

F
NUTRITIONAL IMPROVEMENT ASSUMED EFFECT AT
NATIONAL LEVEL:
J
INDIVIDUAL
DEVELOPMENT EDUCATIONAL

Individual probability Lower

of employment unemployment rate

Access to high income Better qualified

labour force
Lower fertility - - - __- Lower population
increase

Better MCH ------- Lower IMR

.
L
INTERGENERATIONAL NATIONAL
EFFECT DEVELOPMENT

Fig. 4. Assumed or demonstrated associations between nutritional improvement, individual and national
development. IMR, infant mortality rate; MCH, maternal and child health.

Education for women is especially important as it is associated with reduced fertility,

later age of marriage and better child care (Levine, 1980). Educated mothers make
better use of health services and their children have better health care and nutrition and
lower infant mortality rates (Wagner, 1974; Caldwell, 1979). Also better educated
mothers make more effective teachers of their children (Laosa, 1978; Levine, 1980), and
there is some suggestion that their children’s development is also better than children of
less-educated mothers (L,evine, 1980). These may be regarded as specific cases support-
ing the argument put forward by Schultz (1959) for education as an activity economically
justifiable on the grounds that it creates ‘human capital’, with high rates of return
compared with other forms of investment where scarce skills are concerned (Carnoy
et al. 1982).
Almost no studies have been done on the rate of return to the creation of human
capital in Jamaica. In 1980 it was estimated that returns on training in the health care
professions were 20-30% per annum (Cumper, 1982). The general economic conditions
for high rates of return are present in Jamaica; large income differentials between
occupations, occupations linked to educational levels, and a shortage of many skills.
However, limiting the economic importance at national level of nutritional inter-
ventions aimed at supporting education are first, the apparently limited extent of
moderate or severe malnutrition (Samuels, 1987), and second, the small scale of local
high-technology industries. This would not prejudice the importance of such inter-
ventions as a way of increasing equity and improving individual life chances. Also in
countries where malnutrition is more prevalent the effect at a national level would be
expected to be greater.

Downloaded from https://www.cambridge.org/core. IP address: 114.125.72.109, on 23 Dec 2021 at 09:28:48, subject to the Cambridge Core terms of use, available at
https://www.cambridge.org/core/terms. https://doi.org/10.1079/PNS19920012
 78 S. M. GRANTHAM-MCGREGOR AND G . CUMPER

REFERENCES
Caldwell, J. C. (1979). Education as a factor in mortality decline: An examination of Nigerian data. Population
Studies 33, no. 3,395-413.
Carnoy, M., Levin, H., Nugent, R., Sumra, S . , Torres, C. & Unsicker, J. (1982). The political economy of
financing education in developing countries. Financing Educational Development, Proceedings of an
International Seminar, pp. 39-68. Canada: IDRC-205e.
Clarke, N. W., Grantham-McGregor, S. M. & Powell, C. (1991). Health and nutrition predictors of school
failure in Kingston, Jamaica. Ecology of Food and Nutrition 26,41-57.
Cravioto, J. & Arrieta, R. (1979). Stimulation and mental development of malnourished infants. Lancet ii, 899.
Cravioto, J. & Robles, B. (1965). Evolution of adaptive and motor behaviour during rehabilitation from
kwashiorkor. American Journal of Orthopsychology 35,449-464.
Cumper, G. E. (1982). Report of Consultancy on Economic Analysis of the Health Sector in Jamaica: World
Bank Second Population Project.
Grant, J. (1991). The State ofthe World’s Children 1991. Oxford: Oxford University Press.
Grantham-McGregor, S. M. (1982). The relationship between developmental levels and different types of
malnutrition in children. Human Nutrition: Clinical Nutrition 36C, 319-320.
Grantham-McGregor, S. M. (1989). The effects of undernutrition on mental development. In Handbook of the
Psychophysiology of Human Eating, pp. 321-340 [R. Shepherd, editor]. Chichester: J. Wiley and Sons Inc.
Grantham-McGregor, S. M. & Buchanan, E. (1982). The development of an adopted child recovering from
severe malnutrition - A case report. Human Nutrition: Clinical Nutrition 36C, 251-256.
Grantham-McGregor, S. M., Meeks Gardner, J., Walker, S. & Powell, C. (1990). The relationship between
undernutrition activity levels and development in young children. In Activity, Energy Expenditure and
Energy Requirements of Infants and Children, pp. 361-383 [B. Schurch and N. Scrimshaw, editors].
Switzerland: Nest16 Foundation.
Grantham-McGregor, S. M., Powell, C. & Fletcher, P. (1989). Severe malnutrition, stunting and mental
development in young children. European Journal of Clinical Nutrition 43,403409.
Grantham-McGregor, S. M., Powell, C. A., Walker, S. P. & Himes, J. H. (1991a). Nutritional supplemen-
tation, psychosocial stimulation, and mental development of stunted children: the Jamaican Study. Lancet
338, 1-5.
Grantham-McGregor, S. M., Schofield, W. & Powell, C. (1987). Development of severely malnourished
children who received psychosocial stimulation: six year follow-up. Pediatrics 79,247-254.
Grantham-McGregor, S. M., Stewart, M. E. & Desai, P. (1978). A new look at the assessment of mental
development in young children recovering from severe malnutrition. Developmental Medicine and Child
Neurology 20,773-778.
Grantham-McGregor, S. M., Stewart, M. & Powell, C. (1991b). Behaviour observations of severely
malnourished children in a Jamaican hospital. Developmental Medicine and Child Neurology 33, 706-714.
Griffiths, R. (1967). The Ab es of Babies. London: University of London Press.
Lancet editorial (1970). Classification of infantile malnutrition. Lancet u, 302.
Laosa, L. M. (1978). Maternal teaching strategies in Chicano families of varied educational and socioeconomic
levels. Child Development 49, 1129-1135.
Levine, R. (1980). Influences of women’s schooling on maternal behaviour in the Third World. Comparative
Educational Review 24,78-105.
Levinger, B. (1986). School feeding programmes in developing countries: An analysis of actual and potential
impact. AID Evaluation Special Study no. 30. Washington, DC: AID.
Levitsky, D. A. (1979). Malnutrition and hunger to learn. In Malnutrition, Environment and Behavior, pp.
161-179. New York: Cornell University Press.
Lewin, K., Little, A. & Colclough, C. (1982). Adapting to the 1980’s: Taking stock of educational expenditure.
Financing Educational Development. Proceedings of an International Seminar, pp. 13-38. Canada:
IDRC-TS38e.
McLaren, D. S . , Yaktin, V. S . , Kannawati, A. A., Sabbagh, S. & Kadi, Z. (1973). The subsequent mental and
physical development of rehabilitated marasmic infants. Journal of Mental Deficiency Research 17,273-281.
Martorell, R., Mendoza, F. & Castillo, R. (1988). Poverty and stature in children. In Linear Growth
Retardation in Less Developed Countries [J. C . Waterlow, editor]. New York: Raven Press.
Meeks Gardner, J. M., Grantham-McGregor, S. M., Chang, S. M. &Powell, C. A. (1990). Dietary intake and
observed activity of stunted and non-stunted children in Kingston, Jamaica. Part 11: Observed activity.
European Journal of Clinical Nutrition 44,585-593.

Downloaded from https://www.cambridge.org/core. IP address: 114.125.72.109, on 23 Dec 2021 at 09:28:48, subject to the Cambridge Core terms of use, available at
https://www.cambridge.org/core/terms. https://doi.org/10.1079/PNS19920012
 T A S K FORCE O N OVERSEAS M E M B E R S 79

Monckeberg, F. (1990). Treatment of the malnourished child. In The Malnourished Child. Nest16 Nutrition
Workshop Series, pp. 339-358 [R. M. Suskind and L. Lewinter-Suskind, editors], vol. 19. New York:
Raven Press.
Powell, C. A. & Grantham-McGregor, S. M. (1985). Home visiting of varying frequency and child
development. Pediatrics 84, 157-164.
Powell, C., Grantham-McGregor, S. M. & Elston, M. (1983). An evaluation of giving the Jamaican
government school meal to a class of children. Human Nutrition: Clinical Nutrition 37C, 381-388.
Samuels, A. (1987). Health Sector Review 1987: Ministry of Health (Jamaica).
Schiefelbein, E. (1983). Educational Financing in Developing Countries. Ottawa, Canada: IDRC-TS38e.
Schultz, T. W. (1959). Investment in man: an economist’s view. Social Service Review 33, 110-117.
Simeon, D. & Grantham-McGregor, S. M. (1989). The effects of missing breakfast on the cognitive function of
school children of differing nutritional status. American Journal of Clinical Nutrition 49,646-653.
Simeon, D. & Grantham-McGregor, S. M. (1990). Nutritional deficiencies and children’s behaviour and
mental development. Nutrition Research Reviews 3,l-24.
Stevenson, H. W., Parker, T., Wilkinson, A., Bannevaux, B. & Gonzales, M. (1978). Schooling, environment
and cognitive development: A cross cultural study. Monographs of the Society for Research in Child
Development 43, no. 3.
Wagner, D. A. (1974). The development of short term and incidental memory: A cross cultural study. Child
Development 45,389-395.
Winick, M., Meyer, K. K. & Harris, R. C. (1975). Malnutrition and environmental enrichment by early
adoption. Science 190, 1174-1175.

Printed in Great Britain

Downloaded from https://www.cambridge.org/core. IP address: 114.125.72.109, on 23 Dec 2021 at 09:28:48, subject to the Cambridge Core terms of use, available at
https://www.cambridge.org/core/terms. https://doi.org/10.1079/PNS19920012